The Future of Investing

• ETFGI reported global ETF assets above US$24 trillion for the first time at end-August 2026 ($24.03 trillion), and about $23.9 trillion at end-September, up 20.7% year to date.
• Year-to-date net inflows reached a record $2.24 trillion through September, with $276 billion in September alone (ETFGI, Oct 2026).
• iShares, Vanguard and State Street SPDR together managed $13.93 trillion, 58.3% of the total, so scale concentration continues (ETFGI, Oct 2026).

Morningstar's 2026 US Fund Fee Study (May 2026) put the asset-weighted average fund fee at 0.32% for 2025, extending a two-decade decline. Passive funds averaged 0.10% after a 5.4% drop; active fund fees fell 2.7%.

Morningstar's Active/Passive Barometer for the year to June 2026 (published 12 August 2026, covering about $29 trillion, 67% of the US fund market) found just over 40% of active strategies survived and beat passive peers over one year, up 7 points. Over ten years only about 25% did.

Morningstar data cited by PWL Capital (April 2026) show passive funds held over 55% of US fund assets at year-end 2025, up from about 53%, with roughly $951 billion of passive inflows against $187 billion of active outflows. Morningstar dates the passive-over-active crossover to end-2023, earlier than the 2024 framing in Chapter 02.

• Blue Owl closed quarterly redemptions in its non-traded BDC OBDC II from February 2026 and returned about 30% of its NAV through a March capital distribution (Wealth Management, 2026).
• Blackstone's BCRED lifted its Q1 2026 redemption limit from 5% to 7.9% to meet requests; Cliffwater's $33 billion interval fund received requests for 14% of shares against a 7% cap (Wealth Management, 2026).
• Reports put total Q1 2026 private-credit redemption requests above $20 billion; this figure comes from secondary sources and is not independently verified.

This is the illiquidity-versus-liquid-promise tension described in Chapter 05 showing up in practice.

Following the August 2025 executive order, the Department of Labor published a proposed rule on 30 March 2026 (Fiduciary Duties in Selecting Designated Investment Alternatives) creating a safe harbor weighing performance, fees, liquidity, valuation, benchmarks and complexity. Comments closed 1 June 2026; finalisation was reported as targeted for late 2026 (law-firm summaries, 2026). A final rule was not confirmed.

Robinhood announced plans to let customers direct AI agents to trade equities, derivatives, crypto and event contracts (FOW, 28 May 2026). Rollout details were not available to verify, so this Level-4/5 step for retail remains an announcement.

Not refreshed: robo-advised assets and the Vanguard, Schwab, Betterment and Wealthfront figures, direct-indexing projection, private-market asset totals, algorithmic share of volume, retail trading share, SPIVA 15-year figure, Big Three ownership share, Alpha Arena results and the quant-fund figures.

For four centuries the edge in markets belonged to human judgment. In four decades, it moved into software. Systems now sense, decide and execute faster than any person and, increasingly, they learn.

1. 1602First shares: Dutch East India Company shares trade in Amsterdam.
2. 1952Portfolio theory: Markowitz turns risk and return into mathematics.
3. 1975The index fund: Bogle launches Vanguard's first index fund.
4. 1993The ETF: SPY makes the S&P 500 trade like a stock.
5. 2008Robo-advisors: Betterment and Wealthfront automate advice.
6. 2026Agentic AI: LLM agents plan, trade and self-correct.

1. L1 Manual: humans pick and place trades. (Human decides)
2. L2 Algorithmic: pre-set rules execute.
3. L3 Automated: optimised end-to-end.
4. L4 Autonomous: models adapt to new data.
5. L5 Agentic: systems set their own goals. (Machine decides)

The most successful investment strategy of the last fifty years was to stop trying to win. By simply owning the whole market at rock-bottom cost, index funds beat most professionals and took their assets.

1. Lower fees
2. More inflows
3. Greater scale
4. Lower unit costs
5. Cheaper still (and the loop repeats around index economics)

The cost of owning the market fell from about one percent to almost nothing: a price war investors won by doing nothing. When a product is commoditised, price falls to marginal cost. For indexing, that is close to zero.

1. 1975Index the market: Vanguard's mutual structure removes the stock-picker's salary.
2. 1993Wrap it in an ETF: trading efficiency and tax advantages cut costs further.
3. 2018Zero-fee funds: Fidelity launches index funds charging nothing.
4. 2019Zero commissions: Schwab, then all major brokers, drop trading fees.

• Payment for order flow: brokers sell your orders to market makers for a fee.
• Securities lending: your shares are lent to short-sellers; the fund keeps much of the fee.
• Cash sweep: idle cash earns the broker interest, not you.
• Cross-sell: free trading is the funnel for margin loans, advice and premium tiers.

Robo-advisors proved a machine could do an adviser's core job (allocate, rebalance, and harvest losses) for a tenth of the price. Then the incumbents copied them, and the startups' edge became a feature everyone offers.

1. Profile: a questionnaire infers goals, horizon and risk tolerance.
2. Allocate: map the profile to a low-cost ETF portfolio.
3. Rebalance: drift back to target automatically as prices move.
4. Tax-loss harvest: sell losers to bank tax savings, then reinvest.

A two-axis map of automation (horizontal) against personalisation (vertical). Bespoke, hand-made (top left): private banking and human advisor. Bespoke, automated (top right): direct indexing. Middle right: robo-advisor. Lower right: target-date fund and, in the generic, automated corner, the index ETF. Generic, manual (bottom left) is empty.

The best returns migrated to private markets, and the industry is now racing to sell them to everyone else. Fewer companies go public; value is created before the IPO. Access, not allocation, is the new battleground.

• Interval funds: own private assets; redeem a fixed slice each quarter.
• BDCs: listed vehicles that hold portfolios of private credit.
• Semi-liquid funds: Blackstone's BREIT and peers, built for the wealthy.
• 401(k) inclusion: private equity enters US retirement menus.
• Tokenisation: blockchain fractions of funds and assets.
• Evergreen funds: perpetual vehicles with no fixed end date.

The next ambition is not to match the market cheaply, but to beat it with machines that learn. Quant funds proved code can win; now LLMs and agents are reaching for the same prize, in public.

1. Risk and compliance: monitor, stress-test and enforce constraints.
2. Execution: algorithms slice and route orders (e.g. JPMorgan's LOXM).
3. Portfolio construction: optimise positions within risk limits.
4. Signals: models translate information into predicted edge.
5. Research: finance-native LLMs read filings, news and transcripts.
6. Data: real-time and historical, market and non-market signals.

1. Set objectives: the human gives goals and constraints, not trades.
2. Sense and research: agents ingest markets and news continuously.
3. Decide and size: models choose positions and a risk budget.
4. Execute and adapt: trade, monitor outcomes, adapt without reprogramming.

When everyone owns the same index and runs the same model, everyone tries to sell at the same moment. Automation removes human error and adds systemic correlation. Cheaper markets can be more fragile ones.

• Herding: common models crowd the same trades, then unwind together.
• Model homogeneity: if everyone buys the same data, signals converge and diversification fails.
• Flash events: automated selling can drain liquidity in minutes.
• Concentration: three firms vote most of the S&P 500's shares.
• Accountability: you can't jail an algorithm; who owns the harm?
• Regulatory mismatch: over- and under-regulation across jurisdictions.

The genius of the machine age of investing is that it made owning the whole market almost free. Its danger is that when everyone owns the same thing, run by the same code, a market can forget how to disagree, and a market that cannot disagree is one trade away from a stampede.

1. Cheaper won: passive and near-zero fees are permanent, not cyclical.
2. Advice became software: robo and direct indexing commoditised allocation.
3. Access is the next fight: private markets are being opened, carefully or recklessly.
4. Alpha stays scarce: beta democratised; machine alpha did not.
5. Correlation is the new risk: efficiency in the calm, fragility in the crisis.

• Morningstar: US Fund Fee Study; active vs passive flows (2023-24).
• Investment Company Institute (ICI): Fact Book, index fund and ETF data.
• ETFGI: global ETF industry assets (2024).
• S&P Dow Jones Indices: SPIVA Scorecard (active vs index).
• Preqin; McKinsey: Global Private Markets Review (2024).
• BlackRock: Aladdin, iShares and ETF market projections.
• SEC, CFTC, FINRA, BIS: market-structure and algorithmic-trading oversight.
• Company disclosures; Statista; Nof1.ai Alpha Arena. Figures marked 'est.' are compilations.